The study found that most social security offending is committed by a defined population, namely, welfare recipients and, to a lesser extent, claimants. The majority have been entitled to a welfare payment in some amount for some period. Two-thirds of female offenders convicted in the 2 years under study (mid-1989 to mid-1991) were single-parent pensioners. It is likely that most of these women were eligible for the pension when they first received it but became ineligible or received more than their entitlement at a later stage, failing to advise the department of their changed circumstances. About one-half of this defined population of welfare recipients are women. Women are represented in the social security offending population in approximately the same proportion as in the recipient population, and the pensions and benefits misused reflect the different usage by women and men. The study found that magistrates' courts are overwhelmingly more likely to imprison women offenders for this category of offense than for any other category coming before them and that women are more likely to be imprisoned for such offenses than men. The dependent variables that might explain or justify this variance fail to do so. The findings suggest an unduly punitive ethos produced by the combination of professional prosecutors, the practice of making submissions on sentence, inadequate understanding of some of the legal precedents, and a systemic failure to get beneath the surface of offender behavior to the social reality that underlies women's social security offending. Extensive tabular data and a 90-item bibliography